| E-commerce in china has flourished in the new century, the users number of e-commerce has reached hundreds of millions. These means, the traditional structure of e-business data management face more complex application environments , now. And it needs to provide vast amounts of data management, different types of data compatibility, and better user experience. Cloud computing is parallel comput- ing, distributed, grid complex, and it is also one the direction of future development of information technology . It can lower cost provide vast amounts of data processing, compatible with various types of data, and has good scalability.This article is up to study the e-commerce data management by use of cloud computing data management technology. Plan to use cloud computing technology to enhance the current level of e-business data management.The starts is to analysis the current e-commerce Jing dong Mall and Tao bao. Then we identify the characteristics of the two corporate data, and find existing data issues for improvement. Next, we classify these different types data by the Internet information organization method, and propose different solutions. Finally, we have a comprehensive e-business data mana- gement model, which is probably divided into three parts, the user response to part, Hadoop distributed part ,and database systems. Hadoop system part is a open source cloud computing technology. And it is to address the large files in e-commerce systems and unstructured small file storage problems. According to the current situation, we made some adjustments to the native Hadoop system to adapt to our e-commerce environment. Through increasing the main read server share the data flaw with the main system server, and modify the hadoop s file format to improve the efficiency of small file reads. In the database part, this paper presents a distributed master-slave structure, add directory server to get load balancing, and finally made two sub-library strategy. |
PDF Full Text Request